On Premises IIoT Architecture for Capturing, Storing and Visualizing Production Data

In this paper an automated system for production data acquisition, storage and analysis is proposed. It covers the structure and a test of the entire data path consisting of machines, data broker, database and visualization software. Here, unlike other solutions, the actual data source in the production process is variable and not limited to a specific protocol or connector. To achieve this, a locally available gateway running the software Node-RED is used, to receive the data from the different machines and forward it to a central server using MQTT, for example. On this server, the data is stored in the time series database InfluxDB and is consequently available for analysis via Grafana. A particular focus lies on the independent software components, which can be easily exchanged and combined in other ways. Thus, choosing this software means that you won't be affected by the vendor lock-in that is common in the market. Security wise each software component can be hosted on premises, which is a deliberate move away from the cloud solutions that are common in many places. In addition to the structure of the described system, this paper also deals with the performance in terms of writing data to or retrieving data from the database compared to other solutions. Moreover the latency of different approaches in the system is considered.